The Daily Parker

Politics, Weather, Photography, and the Dog

The Cato Institute hopes you can't do math

Some of my libertarian-minded friends have circulated an article written by Cato Institute senior fellow Daniel J. Mitchell, an anti- flat-tax advocate, claiming that Cam Newton will pay a 200% tax to California on his Superbowl earnings. Mitchell quotes "a Certified Public Accountant" writing in a Forbes article at length, ending with this legerdemain:

If the Panthers ... lose [the Superbowl, Newton] will only net another $51,000. The Panthers will have about 206 total duty days during 2016, including the playoffs, preseason, regular season and organized team activities (OTAs)....

Seven of those duty days will be in California for the Super Bowl... To determine what Newton will pay California on his Super Bowl winnings alone ... looking at the seven days Newton will spend in California this week for Super Bowl 50, he will pay the state ... $101,360 on $51,000 should they lose.

Except that's total bullshit. Did anyone else spot the problem with this?

See, Newton didn't earn $51,000 for losing the Superbowl; he earned over $1.1 million for losing the Superbowl. And a $100,000 tax on $1.1 million seems pretty reasonable to me, despite how unreasonable it seems to the Cato Institute (which thinks any tax on income is unreasonable and wants to repeal the 16th Amendment).

If Newton works 206 days in 2016, and 7 of them are in California, then 3.4% of his annual gross income is apportioned to California. But Newton will probably earn $31 million in 2016, not $51,000; and 3.4% of $31 million is, it turns out, $1,053,398.

(Come to think of it, the $51,000 bonus seems kind of small, doesn't it? I mean, since we're talking about fantasy money and not the compensation that most people earn.)

Mitchell's problem isn't that states like California have higher income taxes than other states. His problem is that doesn't want any income taxes, period. Fine; make that argument. But don't foist patently misleading headlines on completely misleading articles and claim you're presenting a real argument.

Bad software testing could have thrown an election

The Daily WTF (a must-read if you're in a technology job) today described how poor testing caused 2,000 ballots to be thrown out in a 2014 election in Brussels:

It wasn’t enough to sway any one election, but the media had already caught wind of the potential voter fraud. Adrien’s company was hired for an independent code review of Delacroy Europe’s voting program to determine if anything criminal had transpired.

He noticed something strange in the UI selection functions, triggered when the user selects a candidate on the viewscreen.

He found two commented lines, dated June 28, 2013, a year before election day. A developer, looking at Card_Unselect(), realized that by unselecting a candidate, it also unselected everyone in that candidate’s list. They commented out two lines, thinking they had fixed the error. However, the unselection algorithm never decremented the check counter, which kept track of how many candidates had been chosen. If a user checked a candidate on one list, changed their mind, and picked another from a separate list, then both votes would be counted.

It hadn’t been a case of fraud, but some poorly-placed comments.

It also could have been prevented—or at least discovered immediately—through automated unit testing.

"Never ascribe to malice what can be adequately explained by incompetence."